The video recordings of our webinar “Quantifying Neighborhood-Level Social Determinants of Health and Risk Landscapes” and workshop “Multivariate Spatial Data Analysis in GeoDa” are now available for viewing and downloading. The October webinar and workshop featured Marynia Kolak, MS, MFA, PhD, Center for Spatial Data Science, University of Chicago. Dr. Kolak is a health geographer and data scientist using open science tools and an exploratory data analytic approach to investigate issues of equity across space and time.


About the Seminar: Associations between social and neighborhood characteristics and health outcomes are well known but remain poorly understood owing to complex, multidimensional factors that vary across geographic space. Growing interest in quantifying social determinants of health (SDOH) at a small-area resolution must account for such complexity. In a recent cross-sectional study, a Kolak-led team developed multidimensional SDOH indices and a regional typology of the continental U.S. at a small-area level using dimension reduction and clustering machine learning techniques, spatializing results at each stage. During a time of increased attention to SDOH, a spatially explicit approach may provide actionable information for key stakeholders with respect to the focus of interventions — and better understand what constitutes, drives, and sustains resilient communities. The webinar covered the results and discussion of this recent cross-sectional study.

About the Workshop: This workshop covered spatial data handling in GeoDa (, exploratory spatial data basics of areal data (e.g. choropleth maps, parallel coordinate plots, etc.), and some advanced multivariate analysis. Participants conducted a principal component analysis and k-means cluster analysis with spatial data, mapping outcomes at each stage for additional exploration and understanding of regional dynamics. The participants used relevant social determinants of health variables for the continental U.S. using 2018 5-year average Census data, and calculated US-wide multidimensional indices and neighborhood typologies at the census tract scale.